X (Twitter) Engagement Rate Benchmarks 2026: What Brands Get Now
Written by: Tim Eisenhauer
Last updated:
What is a good X engagement rate for brands in 2026?
The 2026 industry benchmark median is 0.03% per post, by followers. That’s the good news, because it doubled from 0.015% the year before. It also still rounds to zero at one decimal place. For scale: a brand with 10,000 followers earning the median gets three interactions per post. Hootsuite publishes an average of 1.8% using a formula it doesn’t disclose, which is a 60x disagreement over a platform where, by every follower-based measurement, most industries earn literally nothing. If you want one honest sentence: brand engagement on X is statistically indistinguishable from silence, and the interesting data is about who’s left and who’s leaving.
Our 30-industry benchmarks post called X “dead for most industries” based on the earlier data, and I wondered whether this deep dive would soften that. It didn’t. But the 2026 numbers added two wrinkles I didn’t expect: the median doubled, and the platform’s audience didn’t shrink. X in 2026 is a strange thing: a stable audience of tens of millions of people, on a platform where brands specifically get nothing.
The headline numbers, such as they are
| Source | Overall brand engagement | Formula | Note |
|---|---|---|---|
| Rival IQ 2026 | 0.03% median, doubled from 0.015% | interactions ÷ followers | 150 companies per industry, medians |
| Rival IQ 2025 | fell 48% YoY, worst of any platform | by followers | the collapse that preceded the doubling |
| Hootsuite 2026 | 1.8% average | not stated | the one number both its June articles agree on |
Read that middle row again. X engagement fell 48% in a single year, the worst decline of any platform in the report, and then “doubled” the following year. Doubling off a floor of 0.015% gets you to 0.03%. This is the arithmetic version of a dead-cat bounce, and if a vendor headline next year says “X engagement doubles again,” you now know that could mean brands went from 3 to 6 interactions per 10,000 followers.
By industry, the by-follower medians from the edition our parent post documented: Fashion 0.00%, Food & Beverage 0.00%, Health & Beauty 0.00%, Media 0.01%, Retail 0.01%, Travel 0.01%, Alcohol 0.02%, Nonprofits 0.02%, Tech 0.02%, Financial Services 0.03%, Higher Education 0.04%, Sports Teams 0.07%. Zero, to two decimal places, for three industries. Sports teams “lead” at seven-hundredths of a percent.
Posting more doesn’t fix it
The median brand posts 2.24 times per week on X. Hootsuite’s frequency data says peak engagement (2.21%) comes at exactly 2 posts per week for most industries, and its industry curves make the volume story explicit: healthcare peaks at 28 posts per week and earns 2.65%, barely above the 2.14% dining earns at 8. Fourteen times the posting for half a percentage point. Whatever X’s algorithm rewards, it isn’t brand effort.
There’s one honest exception pattern, and I said it in the parent post: I’m on X constantly, but I follow founders, builders, and people arguing about things, not company accounts. X still works for humans with opinions. It just doesn’t work for logos. We see it with Apaya clients too: when businesses connect their accounts, X is the platform they most often skip entirely.
The audience didn’t leave. The brands did.
Here’s what makes X’s 2026 data genuinely interesting instead of just grim. Pew Research’s late-2025 survey puts X at about 21% of U.S. adults, essentially stable from 22% the year before, with 10% of adults using it daily. DataReportal counts a 99 million U.S. ad audience, about 28.5% of the population. The user base is fine.
Two composition details worth knowing. The politics flipped: 24% of Republicans now report using X versus 19% of Democrats, a reversal from two years prior, which matters if your customer base skews either way. And the milestone almost nobody noticed: Reddit, at 26% of U.S. adults, now reaches more Americans than X does. If your content plan still includes X out of inertia but has never considered where the audience migrated, that’s the stat to sit with.
Meanwhile, the only sector with published presence data is conducting an orderly evacuation. M+R’s 2026 study found just 54% of nonprofits still on X, follower counts down 3%, 31% of those remaining planning to leave, and 65% already building elsewhere. No equivalent study exists for small businesses, which is itself telling: nobody’s even measuring brand presence on X anymore. (The nonprofit benchmarks deep dive has the full exodus data.)
What to do with any of this
Default to skipping it. If you’re a local or SMB brand deciding where posting hours go, the by-follower medians say X returns nothing for eleven of twelve industries. Reassign the time to wherever your customers demonstrably are, and if you want the platform-by-platform comparison, the parent benchmarks post lays it out.
If you stay, stay cheap. Two posts a week is the documented peak of the engagement curve. Anything beyond that is effort the data says the algorithm will not repay. Cross-posting what you already made costs nothing extra when the publishing is automated; bespoke X content is where the waste is.
If X really matters to your business, be a person there. Media, sports, tech, and founder-led accounts are the exceptions with a pulse. A founder posting real opinions will outperform the company account every time, and no scheduling tool changes that.
Watch your own numbers, not the medians. A platform this dead is also this cheap to test: your analytics will tell you within a month whether your audience is among the exceptions. How Apaya publishes there, for the accounts that keep it, is on the X page.
Every number in this post came from a company that sells social media software, and I run one too. The difference is I’m willing to publish the zeros.
What else people ask about X benchmarks
What is the average engagement rate on X?
0.03% median by followers for brand accounts in the 2026 data, up from 0.015%. Hootsuite’s undisclosed-formula average is 1.8%. Individual human accounts routinely do far better than either number; the medians describe logos.
Did X engagement really double in 2026?
Yes, from 0.015% to 0.03%, after falling 48% the year before. Both statements are true, which is a tidy lesson in why percentage-change headlines need the underlying numbers attached.
Is anyone still on X?
About one in five U.S. adults, and one in ten daily. The audience is stable, male-skewing, and now slightly Republican-leaning. The brands are what left, and Reddit now reaches more U.S. adults than X does.
Should nonprofits stay on X?
Their peers are answering: 31% of nonprofits still on X plan to leave, per M+R’s 2026 study, and nonprofit engagement there is 0.02% by followers. See the full nonprofit benchmarks for the sector data.
Sources
- Rival IQ (Quid) — Social Media Industry Benchmark Reports, 2025 and 2026 editions
- Hootsuite — 2026 benchmark articles
- Pew Research Center — Americans’ Social Media Use, November 2025
- DataReportal — Digital 2026: The United States of America
- M+R — Benchmarks 2026 study
Free guide
Scale enterprise social without scaling cost.
See what social content production really costs, how a production system cuts the work, and how to build the case to fund it.
Save 20+ hours a month. Let AI handle your social media.
Apaya writes your posts, designs your graphics, and publishes everywhere — automatically.
